Deconstructing the Testing Mode Effect: Analyzing the Difference Between Writing and No Writing on the Test

Abstract

The examination of the testing mode effect has received increased attention as higher education has shifted to remote testing during the Covid-19 pandemic. We argue that the testing mode effect should be broken into four distinct subparts: the ability to physically write on the test, the method of answer recording, the proctoring/testing environment, and the effect testing mode has on instructor question selection. This paper examines an area largely neglected by the literature surrounding the testing mode effect, the ability (or lack thereof) to write on the test. Using a normalization technique to control for student aptitude and instructor bias, we find that removing the ability of students to physically write on the test significantly lowers student performance. This finding holds across multiple question types classified by difficulty level, Bloom’s taxonomy, and on figure/graph-based questions, and has implications for testing in both face-to-face and online environments.

Keywords

• testing mode
• test environment
• writing on the test
• paper versus computer testing
• test construction